A number of websites including one or plural web pages are connected to the Internet. An information processing apparatus connected to the Internet can provide a wide variety of information items to the user when the user accesses the web pages. The web page corresponds to a file described in Hypertext Markup Language (HTML). The web page is recognized by a uniform resource indicator (URI) or uniform resource locator (URL) (hereinafter referred to as a URI). A website is a collection of web pages. Access to a number of web pages is enabled, and hence it is difficult for a user of an information processing apparatus to find a valuable web page. Thus, processing is known, in which the user of the apparatus designates a remarkable term included in the web pages accessed by the user, on the other hand, the apparatus uses the designated term as a keyword, further extracts the keywords from character strings around the term, searches the web pages by designating the keywords, and accesses the retrieved web pages.
However, in order to cause the apparatus to search the web pages, it has been necessary for the user of the apparatus to select and designate the remarkable term. Accordingly, it has been inevitable that the meaning of the term must be deliberated, and an operation of correctly designating the selected term is required and, as a result of the above, there has been a problem that it is not possible to cause the apparatus to search the web pages easily.
In order to solve the above problem, a system for automatically extracting a keyword from a web page displayed by the information processing apparatus has been developed. The extracted keyword is displayed on the screen, and hence the user can carry out keyword searching by only selecting the displayed keyword.
Although a conventional information processing apparatus has been able to select any keyword from among keywords extracted from the web page currently displayed by the apparatus, the information processing apparatus has not been able to select a keyword from keywords extracted from a web page displayed in the past.